CN105357678A - Random network calculation based wireless sensor network energy allocation and evaluation method - Google Patents

Abstract

The invention provides a random network calculation based wireless sensor network energy allocation and evaluation method, relating to the field of wireless sensor energy management, and solving the problem of power supply deployment of wireless sensor network nodes powered by renewable energy. System parameters of wireless sensor network nodes, comprising power supply, power consumption and power demand, are recorded; according to the cumulant of recorded power supply and the power consumption, the lower limit of power supply level, the upper limit of the power supply, the upper limit of power demand and the lower limit of the power demand for normal working of wireless sensor network nodes are acquired; according to the lower limit of the power supply, the upper limit of the power supply, the upper limit of the power demand and the lower limit of the power demand, the power undersupply probability and power supply waste probability are acquired; and a proper battery capacity is selected for evaluation and adjustment of the power supply deployment of the wireless sensor network nodes. The method provides a basic guide for power generation and storage system configuration of the renewable energy of the wireless sensor network.

Description

The random network calculation method of wireless sensor network energy source configuration and evaluation

Technical field

The present invention relates to wireless sensor energy management domain, be specifically related to a kind of wireless sensor network regenerative resource configuration based on random network calculation and evaluation method.

Background technology

Wireless sensor network is the sensor network of a new generation, the eurypalynous transducer of the crowd had, detectablely comprises diversified phenomenon in earthquake, electromagnetism, temperature, humidity, noise, luminous intensity, pressure, soil constituent, the size of mobile object, the surrounding enviroment such as speed and direction.The application and development of wireless senser brings far-reaching influence for human lives, comprises each productive life fields such as military affairs, aviation, anti-terrorism, explosion-proof, the disaster relief, environment, medical treatment, health care, household, industry, business.But limit by subordinate's mode and operational environment, sensor node is generally difficult to obtain by power circuit the supply of electric power continued.The Power supply problem of wireless sensor network directly affects its life span and service behaviour, is one of guardian technique limiting its job applications.How to develop important topic urgently to be resolved hurrily for wireless sensor network node provides the supply of electric power effectively continued to become wireless sensor network.

In recent years, the important directions of wireless sensor network application study has been become based on the node electric power system of the regenerative resources such as solar energy.The main power consumption of wireless sensor network node is by biosensor power consumption, and data transmit-receive power consumption, power consumption of processing unit and standby dormancy power consumption form.As ZigBee wireless sensor network node conventional at present, the operating voltage of 3.3V.Peak current during data send is 29mA, and the peak current during data receiver is 24mA, and the operating current of various transducer is about 30mA.

Suppose employing 8051 kernel, operating current is 6mA, and resting state is about 1 μ A.The power consumption then performing once-through operation processor is 19.8mW.During data send, power consumption is 115.5mW, and during data receiver, power consumption is 99mW, and data acquisition period power consumption is 118.8mW, and all operations is concurrent, and power consumption is no more than 293.7mW, about 300mW.Lithium battery normal working voltage scope 3.7-4.2V, if capacity chosen by battery is 900mAh, then can power continuously 12 hours when continuous peak value concurrent efforts, if battery selects capacity to choose 3000mAh, then can power continuously 40 hours under same case.Consider that the region of radio node subordinate may be positioned at south China, when continuous overcast and rainy season also likely ten days not sunny situations, then choosing of battery capacity should be larger.But actual conditions are wireless sensor network nodes is generally timing acquiring to the collection of data, as temperature, the data that humidity changes at a slow speed.Such as work 50s per hour, other times are all in dormancy, and electric current is with microammeter, and power consumption is negligible.So the battery capacity calculating gained in above-described continuous peak value concurrent efforts situation is just too strict, and cause battery capacity to be chosen excessive, node volume strengthens, and manufacturing cost increases.

Network calculus proposes based on queuing theory in recent years, network service carried out to the advanced technology means of qualitative and quantitative analysis.Its application boundary describes arrival in queue system and service characteristic, and assesses queue performance.Network calculus almost can be applied to the various aspects of computer and communication network, and can expand to other various productive life fields, as traffic, and air net etc.The key idea of network calculus a complicated non-linear queue system is converted to suitable (min ,+) algebraically that one is easy to treatment and analysis.The development of network calculus is divided into two directions that are mutually related: certainty calculation and randomness calculation.Deterministic network calculus can derive the border of compacting.But these borders are under the system configuration of high complexity, or may seem that some is too conservative when the situation that a tolerable small probability is run counter to.This just excites the development of randomness network calculus, and it has many-sided Statistical Superiority, thus obtains the valid conclusion of resource capacity problem.

Summary of the invention

The object of the invention is to solve and to be powered deployment issue by the wireless sensor network node of regenerative resource power supply, while the electric power supply that effective guarantee wireless network node is continual and steady, make full use of battery electric quantity and reduce energy waste.

Based on the configuration of wireless sensor network regenerative resource and the evaluation method of random network calculation, the method is realized by following steps:

Step one, record wireless sensor network node system parameters, comprise electric power supply C (t), power consumption C ^*(t) and electrical energy demands S (t); And the cumulant C (s, t) of electric power supply calculated in time interval [s, t] and the cumulant S (s, t) of electrical energy demands, be expressed as with following formula respectively:

C(s，t)＝C(t)-C(s)

S(s，t)＝S(t)-S(s)

In above formula, C (s) is the electric power supply in s moment, s and t is the time, and s is more than or equal to 0 and is less than or equal to t;

Step 2, according to the cumulant C (s, t) of electric power supply obtained in step one and power consumption C ^*t (), adopts following formula respectively: obtain the lower boundary maintaining the normal work energy supply of wireless sensor network node;

In formula, α ₁for the lower bound curve of electric power supply, f ₁for electric power supply lower boundary;

The coboundary of wireless sensor network node electric power supply;

In formula, α ₂for the upper bound curve of electric power supply, f ₂for electric power supply coboundary,

The coboundary of wireless sensor network node electrical energy demands;

In formula, β ₁for the upper bound curve of electrical energy demands, g ₁for the coboundary of electrical energy demands;

The lower boundary of wireless sensor network node electrical energy demands;

In formula, β ₂for the lower bound curve of electrical energy demands, g ₂for the lower boundary of electrical energy demands;

Step 3, lower bound curve α according to the electric power supply obtained in step 2 ₁, electric power supply lower boundary f ₁, electrical energy demands upper bound curve β ₁, electrical energy demands coboundary g ₁, electric power supply upper bound curve α ₂, electric power supply coboundary f ₂, electrical energy demands lower bound curve β ₂and the lower boundary g of electrical energy demands ₂, adopt following formula respectively, obtain electric power supply shortfall probability and electric power supply waste probability;

In formula, the quantity that L (t) lacks for t electric power supply, B is battery capacity, and W (t) exceeds the quantity that battery capacity is wasted for t electric power supply amount;

Step 4, the electric power supply shortfall probability obtained according to step 3 and electric power supply waste probability choose suitable battery capacity, assess wireless sensor network node deployment of powering and adjust.

In the present invention, described electric power supply lower boundary f ₁, electric power supply lower bound curve α ₁, with following formula tabular form be:

$f_1\left(x\right)=\frac{e^{-{\mathrm{\θ\θ}}_1}}{1-e^{{\mathrm{\θ\θ}}_1}}{e}^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0,{\mathrm{\θ}}_1>0$

α ₁＝[ρ(θ，t)+θ ₁]t

$p(\mathrm{\θ},t)\≥\frac{1}{\mathrm{\θ}t}\log E\[e^{\mathrm{\θ}C\left(t\right)}\]$

Electric power supply coboundary f ₂with the upper bound curve α of electric power supply ₂be expressed as with following formula:

$f_2\left(x\right)=\frac{e^{-{\mathrm{\θ\θ}}_2}}{1-e^{-{\mathrm{\θ\θ}}_2}}{e}^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0,{\mathrm{\θ}}_2>0$

α ₂＝[ρ(θ，t)+θ ₂]t

$\mathrm{\ρ}(\mathrm{\θ},t)\≥\frac{1}{\mathrm{\θ}t}\log E\[e^{\mathrm{\θ}C\left(t\right)}\]$

The coboundary g of electrical energy demands ₁with the upper bound curve β of electrical energy demands ₁be formulated as:

g ₁(x)＝e ^-θx，

β ₁＝μ(θ)t+g ₁(x)

$\mathrm{\μ}\left(\mathrm{\θ}\right)=\frac{1}{-\mathrm{\θ}(t-t_0)}\log E\[e^{-\mathrm{\θ}S(t_0,t)}\]$

The lower boundary g of described electrical energy demands ₂with the lower bound curve β of electrical energy demands ₂be formulated as:

g ₂(x)＝e ^-θx，

β ₂＝μ(θ)t-g ₂(x)

$\mathrm{\μ}\left(\mathrm{\θ}\right)=\frac{1}{-\mathrm{\θ}(t-t_0)}\log E\[e^{-\mathrm{\θ}S(t_0,t)}\]$

In above formula, θ, θ ₁and θ ₂for time constant, x is probability parameter.

Beneficial effect of the present invention: the present invention proposes the configuration of a kind of wireless sensor network regenerative resource based on random network Calculus Theory and evaluation method, avoids the waste that conventional arrangement method may cause in Logistics networks node power under continuing the prerequisite of supplying.Random network Calculus Theory combines with wireless sensor network node energy management by the method, establish a wireless sensor network random supply network Calculus Theory framework, effective Performance Evaluation is carried out to the specification of renewable energy conversion equipment (as solar panels) and memory device (as battery) and choosing of capacity.The present invention gives clear and definite electric power supply fail-safe analysis index: electric power supply shortfall probability and electric power supply waste probability, for wireless sensor network regenerative resource electrical energy production and storage system configuration provide the foundation guidance.

Accompanying drawing explanation

Fig. 1 is 140 × 140mm in the random network calculation method of wireless sensor network energy source configuration of the present invention and evaluation ²the electric energy schematic diagram that mono-crystalline silicon solar plate provided in 30 days;

Fig. 2 is 140 × 140mm ²the Electric energy accumulation schematic diagram that mono-crystalline silicon solar plate provided in 30 days;

Fig. 3 is 140 × 140mm ²electric energy accumulation in the mono-crystalline silicon solar plate odd-numbered day arrives up-and-down boundary schematic diagram;

Fig. 4 is the energy ezpenditure schematic diagram in the ZigBee wireless sensor network node odd-numbered day;

Fig. 5 is the up-and-down boundary schematic diagram of the power consumption requirements in the ZigBee wireless sensor network node odd-numbered day;

Fig. 6 is electric power supply deficiency and waste probability schematic diagram in the random network calculation method of wireless sensor network energy source configuration of the present invention and evaluation.

Embodiment

Embodiment one, composition graphs 1 to Fig. 6 illustrate present embodiment, the random network calculation method of wireless sensor network energy source configuration and evaluation, and the method is realized by following steps:

Step one, acquisition wireless sensor network node system parameters: composition graphs 1 to Fig. 3, sets forth one piece of 140 × 140mm ²the power levels that mono-crystalline silicon solar plate provided in 30 days; Electric energy accumulation amount in 30 days and ZigBee wireless sensor network node Energy Expenditure Levels in the odd-numbered day;

Setting Electric energy accumulation, i.e. electric power supply C (t); Power consumption C ^*(t); The service that system provides, i.e. electricity needs S (t).

C(s，t)＝C(t)-C(s)(1)

S(s，t)＝S(t)-S(s)(2)

C (s, t) is the electric power supply cumulant between time interval [s, t], S (s, the t) service that system provides between time interval [s, t], i.e. electrical energy demands cumulant; For electrical energy supply system, be the electric energy that electrical energy demands provides;

Step 2, utilization (3) formula represent the lower boundary providing wireless sensor network node normally to work power supply level:

α ₁represent the lower bound curve of supply of electric power, the lower boundary that namely data arrive in random network calculation.F ₁represent supply of electric power lower boundary, infimum is asked in inf [] expression.

$f_1\left(x\right)=\frac{e^{-{\mathrm{\θ\θ}}_1}}{1-e^{{\mathrm{\θ\θ}}_1}}{e}^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0,{\mathrm{\θ}}_1>0$

α ₁＝[ρ(θ，t)+θ ₁]t

$\mathrm{\ρ}(\mathrm{\θ},t)\≥\frac{1}{\mathrm{\θ}t}\log E\[e^{\mathrm{\θ}C\left(t\right)}\]$

Described time constant and probability parameter are chosen for respectively: θ=0.1, θ ₁=10, x=0.2, E [] are mathematic expectaion,

(4) formula of utilization represents the coboundary of the total charge volume of wireless sensor network node battery, i.e. the coboundary of supply of electric power:

α ₂represent the upper bound curve of supply of electric power, the coboundary that namely data arrive in random network calculation.F ₂represent electric power supply coboundary, supremum is asked in sup [] expression.

$f_2\left(x\right)=\frac{e^{-{\mathrm{\θ\θ}}_2}}{1-e^{-{\mathrm{\θ\θ}}_2}}{e}^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0,{\mathrm{\θ}}_2>0$

α ₂＝[ρ(θ，t)+θ ₂]t

$\mathrm{\ρ}(\mathrm{\θ},t)\≥\frac{1}{\mathrm{\θ}t}\log E\[e^{\mathrm{\θ}C\left(t\right)}\]$

In present embodiment, time constant and probability parameter are chosen for respectively: θ=0.1, θ ₂=10, x=0.2.E [] is mathematic expectaion, with odd-numbered day data instance, and step 2, the up-and-down boundary curve α of the electric power supply that step 3 draws ₁and α ₂provide in the diagram.

(5) formula of utilization represents the coboundary of wireless sensor network node electric energy:

Wherein β ₁represent the upper bound curve of electrical energy demands, i.e. the coboundary of data, services in random network calculation.G ₁represent electrical energy demands coboundary.

g ₁(x)＝e ^-θx，

β1＝μ(θ)t+g ₁(x)

$\mathrm{\μ}\left(\mathrm{\θ}\right)=\frac{1}{-\mathrm{\θ}(t-t_0)}\log E\[e^{-\mathrm{\θ}S(t_0,t)}\]$

In present embodiment, time constant and probability parameter are chosen for respectively: θ=0.1, x=0.2.E [] is mathematic expectaion.

(6) formula of utilization represents that wireless sensor network node electric energy estimates the lower boundary consumed:

Wherein β ₂represent the lower bound curve of electrical energy demands, i.e. the lower boundary of data, services in random network calculation.G ₂represent electrical energy demands lower boundary.

g ₂(x)＝e ^-θx，

β ₂＝μ(θ)t-g ₂(x)

$\mathrm{\μ}\left(\mathrm{\θ}\right)=\frac{1}{-\mathrm{\θ}(t-t_0)}\log E\[e^{-\mathrm{\θ}S(t_0,t)}\]$

In present embodiment, time constant and probability parameter are chosen for respectively: θ=0.1, x=0.2.E [] is mathematic expectaion.

With odd-numbered day data instance, the up-and-down boundary curve β of the electrical energy demands obtained in step 2 ₁and β ₂provide in Figure 5.

Step 3, utilization (7) formula ask for supply of electric power shortfall probability:

l (t) represents that the quantity that t electric power supply lacks, B are battery capacity.Wherein convolution operation is added for minimum.

$g_1\⊗f_1\left(X\right)={\inf }_{0\≤x\≤X}\{g_1\left(x\right)+f_1(X-x)\}$

(8) formula of utilization asks for electric power supply waste probability:

wherein W (t) represents the quantity that t electric power supply amount exceeds battery capacity and is wasted. convolution operation is added for minimum

$f_2\⊗g_2\left(X\right)={\inf }_{0\≤x\≤X}\{f_2\left(x\right)+g_2(X-x)\}$

Above formula is variable is the minimum computational methods adding convolution of X, and x changes between [0, X].

With odd-numbered day data instance, when selected battery capacity is increased to 300 from 100, the electric power supply shortfall probability that step 3 draws and electric power supply are wasted probability and are provided in figure 6.The visible increase along with battery capacity, electric power supply shortfall probability and electric power supply waste probability exponentially level successively decrease.

Step 4, according to the system parameters asked for above, and corresponding environmental parameter, assesses wireless sensor network node deployment of powering and adjusts, choose suitable battery capacity, ensure the work of wireless sensor network continuous effective.

Claims (2)

1. the random network calculation method of wireless sensor network energy source configuration and evaluation, it is characterized in that, the method is realized by following steps:

Step one, record wireless sensor network node system parameters, comprise electric power supply C (t), power consumption C ^*(t) and electrical energy demands S (t); And the cumulant C (s, t) of electric power supply calculated in time interval [s, t] and the cumulant S (s, t) of electrical energy demands, be expressed as with following formula respectively:

C(s，t)＝C(t)-C(s)

S(s，t)＝S(t)-S(s)

In above formula, C (s) is the electric power supply in s moment, s and t is the time, and s is more than or equal to 0 and is less than or equal to t;

Step 2, according to the cumulant C (s, t) of electric power supply obtained in step one and power consumption C ^*t (), adopts following formula respectively: obtain the lower boundary maintaining the normal work energy supply of wireless sensor network node;

In formula, α ₁for the lower bound curve of electric power supply, f ₁for electric power supply lower boundary;

The coboundary of wireless sensor network node electric power supply;

In formula, α ₂for the upper bound curve of electric power supply, f ₂for electric power supply coboundary,

The coboundary of wireless sensor network node electrical energy demands;

In formula, β ₁for the upper bound curve of electrical energy demands, g ₁for the coboundary of electrical energy demands;

The lower boundary of wireless sensor network node electrical energy demands;

In formula, β ₂for the lower bound curve of electrical energy demands, g ₂for the lower boundary of electrical energy demands;

Step 3, lower bound curve α according to the electric power supply obtained in step 2 ₁, electric power supply lower boundary f ₁, electrical energy demands upper bound curve β ₁, electrical energy demands coboundary g ₁, electric power supply upper bound curve α ₂, electric power supply coboundary f ₂, electrical energy demands lower bound curve β ₂and the lower boundary g of electrical energy demands ₂, adopt following formula respectively, obtain electric power supply shortfall probability and electric power supply waste probability;

In formula, the quantity that L (t) lacks for t electric power supply, B is battery capacity, and W (t) exceeds the quantity that battery capacity is wasted for t electric power supply amount;

Step 4, the electric power supply shortfall probability obtained according to step 3 and electric power supply waste probability choose suitable battery capacity, assess wireless sensor network node deployment of powering and adjust.

2. the random network calculation method of wireless sensor network energy source configuration according to claim 1 and evaluation, is characterized in that, described electric power supply lower boundary f ₁, electric power supply lower bound curve α ₁, with following formula tabular form be:

$f_1\left(x\right)=\frac{e^{-{\mathrm{\θ\θ}}_1}}{1-e^{-{\mathrm{\θ\θ}}_1}}{e}^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0,{\mathrm{\θ}}_1>0$

α ₁＝[ρ(θ，t)+θ ₁]t

$\mathrm{\ρ}(\mathrm{\θ},t)\≥\frac{1}{\mathrm{\θ}t}\log E\[e^{\mathrm{\θ}C\left(t\right)}\]$

Electric power supply coboundary f ₂with the upper bound curve α of electric power supply ₂be expressed as with following formula:

$f_2\left(x\right)=\frac{e^{-{\mathrm{\θ\θ}}_2}}{1-e^{-{\mathrm{\θ\θ}}_2}}{e}^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0,{\mathrm{\θ}}_2>0$

α ₂＝[ρ(θ，t)+θ ₂]t

$\mathrm{\ρ}(\mathrm{\θ},t)\≥\frac{1}{\mathrm{\θ}t}\log E\[e^{\mathrm{\θ}C\left(t\right)}\]$

The coboundary g of electrical energy demands ₁with the upper bound curve β of electrical energy demands ₁be formulated as:

$g_1\left(x\right)=e^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0$

β ₁＝μ(θ)t+g ₁(x)

$\mathrm{\μ}\left(\mathrm{\θ}\right)=\frac{1}{-\mathrm{\θ}\left(t-t_0\right)}\log E\[e^{-\mathrm{\θ}S\left(t_0,t\right)}\]$

The lower boundary g of described electrical energy demands ₂with the lower bound curve β of electrical energy demands ₂be formulated as:

$g_2\left(x\right)=e^{-\mathrm{\θ}x},\∀\mathrm{\θ}>0$

β ₂＝μ(θ)t-g ₂(x)

$\mathrm{\μ}\left(\mathrm{\θ}\right)=\frac{1}{-\mathrm{\θ}\left(t-t_0\right)}\log E\[e^{-\mathrm{\θ}S\left(t_0,t\right)}\]$

In above formula, θ, θ ₁and θ ₂for time constant, x is probability parameter.